mirror of
https://git.squeakyrat.moe/faggot-chan/
synced 2026-11-04 18:35:45 +01:00
106 lines
4.4 KiB
Markdown
106 lines
4.4 KiB
Markdown
# how to run ts bullshit
|
|
|
|
two parts: `bot.py` talks to discord, `main.py` (ollama) or `gemini.py` (gemini api) is the brain.
|
|
`main.py` is stateless so it can live on another machine / colab and talk to `bot.py` through a websocket or tunnel, you never have to upload your files anywhere sketchy
|
|
|
|
# installing
|
|
pick your distro. all of them use a venv because python doesn't let you `pip install` system wide anymore
|
|
|
|
### ubuntu
|
|
```console
|
|
sudo apt install python3 python3-venv curl zstd
|
|
curl -fsSL https://ollama.com/install.sh | sh
|
|
```
|
|
### arch
|
|
```console
|
|
sudo pacman -S python ollama
|
|
#ollama-cuda or ollama-rocm if you have a gpu
|
|
```
|
|
### gentoo
|
|
```console
|
|
sudo emerge --ask dev-lang/python sci-ml/ollama-bin
|
|
#ollama-bin is from guru iirc you know how to do that you use gentoo
|
|
```
|
|
|
|
(skip the ollama part if you're using the gemini api)
|
|
|
|
then your python dependencies are
|
|
```console
|
|
python3 -m venv venv
|
|
source venv/bin/activate
|
|
pip install ollama websockets discord.py
|
|
#and this one only if you're using the gemini api, you don't need ollama if so
|
|
pip install google-genai
|
|
```
|
|
|
|
# once you install your dependencies
|
|
you need ollama running (not needed for gemini)
|
|
```console
|
|
#ollama needs to be serving, the packages above already ship a service
|
|
sudo systemctl enable --now ollama #ubuntu / arch / gentoo systemd
|
|
sudo rc-update add ollama default && sudo rc-service ollama start #gentoo openrc
|
|
#or just run it by hand / use tmux or screen
|
|
ollama serve &
|
|
ollama pull <your model>
|
|
ollama pull <your embedding model>
|
|
#those two are in main.py
|
|
#test it with
|
|
ollama run <your model>
|
|
```
|
|
if that runs fine (or you have a gemini key) then you can
|
|
- go to the discord developer portal and turn on **message content intent** and **server members intent** for your bot or it won't start
|
|
- add the bot to your server
|
|
- run the wizard, it writes all the config files for you using your input
|
|
```console
|
|
python3 setup_wizard.py
|
|
```
|
|
- edit your system prompt in `./files/config/system_prompt.txt` (it makes a placeholder the first time bot.py runs)
|
|
- optional: `./config/params.json` has the ollama options (temperature and stuff), without it ollama just uses its defaults
|
|
- run main first, then the bot
|
|
```console
|
|
python3 main.py #or gemini.py
|
|
python3 bot.py
|
|
```
|
|
she only reads `general` and `announcements`, and every dm gets its own memory in `./files/log/dm/` so it never leaks into the server
|
|
|
|
# fallback
|
|
if you're using gemini you can have a local ollama `main.py` take over when gemini says you're out of quota. the wizard asks about it (it's the third line of `./files/config/connection.txt`), it needs to run on another port so run it as
|
|
```console
|
|
PORT=8766 python3 main.py
|
|
```
|
|
the bot goes back to trying gemini after an hour. if there's no fallback (or it's down) she just says "usage limit reached, see you in a month"
|
|
|
|
# admin
|
|
`admin.py` is the admin panel, it only edits files on the box so nobody on discord can touch it. the bot re-reads everything on every message, no restart needed
|
|
```console
|
|
python3 admin.py list
|
|
python3 admin.py edit <blocked|prompt|settings|reactions|params> #opens $EDITOR, vim if you don't have one
|
|
python3 admin.py block <user id>
|
|
python3 admin.py unblock <user id>
|
|
```
|
|
`reactions.json` is where the stickers / emojis / media she can send live, the bot adds new ones with an empty description and you fill them in. media with no description never gets offered
|
|
|
|
you can let the local gemma-4 write the descriptions for you, it needs `pip install pillow` (and ffmpeg for videos), ollama running, and the model has to be able to see images
|
|
```console
|
|
python3 describe.py
|
|
#only the empty ones, --all redoes everything, --model <name> to use another model
|
|
python3 --gemini uses the gemini api instead so you don't need ollama
|
|
```
|
|
|
|
# colab
|
|
if your pc can't run the model you can host `main.py` on colab (or a vps with a gpu), run `setup_wizard.py` on both sides and say yes to the colab question
|
|
```console
|
|
#on colab
|
|
python3 colab_setup.py
|
|
```
|
|
it prints a wss:// address and a token, paste those into the wizard on the bot.py side. the address changes every time you restart the notebook
|
|
this script assumes collab but it probably works with any ubuntu vps. it's not reccomended to run it on anything that's not google colab
|
|
|
|
# IF IT FAILS
|
|
- check `./files/log/errorlog.txt`
|
|
- `bad auth token` means the token in `./files/config/connection.txt` isn't the one main.py has in `./config/auth_token.txt`
|
|
- on colab main.py logs to `main.log` next to colab_setup.py
|
|
|
|
# have fun
|
|
don't ask me why the bot is female she just is
|